Concentration, load and toxicity of spraydrift-borne azinphos-methyl at the inlet and outlet of a constructed wetland

Constructed wetlands have been widely used to treat different types of agricultural point and nonpoint source pollution including nutrients, sediments and some herbicides. However, very little is known about the effectiveness of constructed wetlands in controlling nonpoint source insecticide pollution. Here we describe the concentrations of total suspended solids (TSS), nutrients and spraydrift-borne azinphos-methyl (AZP) at the inlet and outlet of a 0.44-ha vegetated pond. One tributary, which receives contaminants from a 15-ha orchard area, flows directly through the wetland, before it enters the Lourens River, South Africa. A tracer experiment with Rhodamin B was undertaken to find the optimal timing for inlet and outlet sampling performance. The retentions of TSS, ortho -phosphate and nitrate during dry weather conditions were 15, 54 and 70%. Concentration levels of AZP were reduced below the wetland by 90.8 +/- 0.7%, from 0.65 +/- 0.08 to 0.06 +/- 0.01 mug/l (n = 5). The reduction of AZP load was 54.1 +/- 3.8%. A 24-h in situ bioassay employing midge larvae (Chironomus spec.) revealed a significant (ANOVA, Fisher's PLSD; P less than or equal to 0.01) reduction of toxicity. Mortality in two separate trials was 41.3 +/- 2.4% and 46.3 +/- 1.2% at the inlet station and 2.5 +/- 1.4 and 3.8 +/- 1.2% at the outlet station during days with spraydrift deposition into the tributary upstream of the wetland. During time intervals without any spraying in the catchment, mortality of the midge larvae in both trials was 1.2 +/- 1.2% at the inlet station and zero at the outlet station (n = 4 for each trial). We conclude that construction of small vegetated ponds is a suitable risk reduction strategy for agricultural nonpoint source insecticide input into surface waters. (C) 2001 Elsevier Science B.V. All rights reserved.